The human immunodeficiency virus, (HIV), is believed to be the causative agent of Acquired Immunodeficiency Syndrome (AIDS). The nucleic acid sequence of the HIV proviral genome has been deduced and the location of various protein coding regions within the viral genome has been determined.
Of particular interest to the present invention is that portion of the HIV genome known in the art as the gag region. The gag region is believed to encode a precursor protein that is cleaved and processed into three mature proteins, p17, p24 and p15. The HIV p24 protein has an apparent relative molecular weight of about 24,000 daltons and is known in the art as the HIV core antigen because it forms the viral capsid.
The p24 antigen of HIV is of particular interest because studies have indicated that the first evidence of anti-HIV antibody formation (sero- conversion) in infected individuals is the appearance of antibodies induced by the p24 antigen, i.e., anti-p24 antibodies. In addition, recent studies have reported that p24 protein can be detected in blood samples even before the detection of anti p24 antibodies. Detecting the presence of either the p24 protein or anti-p24 antibodies therefore appear to be the best approach to detecting HIV infection at the earliest point in time.
Furthermore, the p24 antigen reappears in the blood of infected individuals concomitant with the decline of anti-p24 antibody in patients showing the deterioration in their clinical condition that accompanies transition into full-blown AIDS. Thus, the p24 antigen can serve as an effective prognostic marker in patients undergoing therapy.
The development of immunoassays for detecting anti-p24 antibodies has been limited by difficulties in producing sufficient quantities of HIV p24 protein that is essentially free of immunoreactive contaminants. The presence of contaminants that immunoreact with antibodies present in patient samples results in lower assay specificity and sensitivity and an increase in false positive results.
Presently, when assaying for anti-p24 antibodies in a blood sample, the art typically overcomes the presence of contaminants in HIV p24 protein preparations by preabsorbing the sample with a preparation containing the contaminants. For instance, Dowbenko et al., Proc Natl Acad Sci USA 82: 7748-7752 (1985) reported using recombinant DNA methods to produce in E. coli a HIV p24 fusion protein. However, an immunoaffinity-purified preparation of the p24 fusion protein contained a level of E. coli protein contaminants sufficient to require preabsorbing the blood samples being tested with E. coli protein extracts.
Similarly, Steimer et al., Virol., 150: 283-290 (1986) reported producing a truncated HIV p24 in E. coli. The truncated p24 was isolated from contaminating E. coli proteins using ammonium sulfate precipitation and fractionation by gel filtration to produce a p24 antigen preparation described as being greater than 99% pure. However, use of that p24 antigen preparation in an enzyme linked immunosorbent assay (ELISA) to detect anti-p24 antibodies still required preadsorbing the blood samples with E. coli proteins. European patent application No. 85309454.8, published Jul. 9, 1986 (publication No. 0187041) also describes the expression of HIV p24 fusion protein in E. coli.
Two difficulties in using genetically engineered E. coli to produce a HIV p24 antigen preparation that is essentially free of contaminating E. coli proteins are insolubility and low yield of the recombinantly produced protein. For example, Dowbenko et al., supra, Ghrayeb et al., DNA, 5: 93-99 (1986) and Shoeman et al., Anal. Biochem., 161: 370-379 (1987) have reported that HIV p24 fusion proteins produced in E. coli accumulated as insoluble aggregates ("inclusion bodies") within the producing bacteria. The aggregates, which can be seen as granules in electron micrographs of the bacteria, were recovered in the pellet fraction after cell lysis (breakage) and centrifugation. Solubilization of the protein from the pellet then required treatment with strong denaturing agents.
The yield of a recombinant protein from transformed bacteria is directly related to its translation rate. A recombinant protein cannot be synthesized faster than the slowest step in the entire translation process. Translation initiation, elongation and termination have all been found to be steps whose efficiency is not predictable from DNA sequence alone.
For example, Steimer et al., supra, reported examining the effect on translation efficiency of varying the three nucleotides 5' of the initiator AUG in a rDNA coding for an amino-terminal truncated HIV p24 protein. This area of the rDNA was examined because it is involved in defining the translation initiation (ribosome binding) site. Steimer et al. found that the efficiency of translation initiation depended upon the integration of several factors and was not predictable from the DNA sequence of the ribosome binding site region.